Salmon Escapement Assessment in the Toklat River, 1995 and 1996

Barton, Louis

In 1995 and 1996 two user non-configurable, sonar salmon counters were operated from opposite banks to estimate salmon passage in the Toklat River upstream of Barton Creek from·~ approximately mid-August through early October. Estimated passage was 110,867 fish in 1995 and 90,044 fish in 1996. The median day of passage was 14 September in 1995 and 16 September 1996. Sonar counting range was considered adequate for the detection of the majority of fish passing the sonar site each year as most were near-shore oriented, passing within 3-4 m of the shoreline. However, passage was greatest along the left bank in both years, representing 81% of the estimate in 1995 and 66% in 1996. Daily passage was greatest during periods of darkness in both years, with the greatest movement occurring on the average between 2100 and 0I00 hours. Apportionment of sonar counts was based upon species composition observed during subsequent ground surveys made of the major spawning area at Toklat Springs in mid-October each year. Approximately, 99% of the sonar estimate in both years was apportioned to fall-run'churri salmon Oncorhynchus Ireta (110,201 in 1995; 88,513 in 1996), with the remainder considered to be coho salmon 0. kisutch The resulting estimates of chum salmon were substantially greater than subsequent ground survey estimates made at Toklat Springs each year. The Toklat Springs total abundance estimate in 1995 was 54,513 chum salmon and revealed the escapement goal was achieved in that year. However, the total abundance estimate in 1996 was only 18,264 chum salmon, well below the minimum escapement goal of 33,000 fish. Although an additional 1,293 chum and 194 coho salmon passed Barton Creek weir from mid-August through early October in 1995, only a single chum and no coho salmon passed the weir in 1996. Variations in Toklat River water levels and velocities, together with migration behavior of upstream migrant salmon affected the ability of the hydroacoustic equipment to accurately estimate salmon passage in both 1995 and 1996: Positive and negative biases in fish passage from sonar counter ping-rate settings were accounted for by comparing sonar counter output to visual observations on an oscilloscope. However, multiple counts as a result of salmon swimming upstream through the acoustic beam more than once was considered to have been the greatest non-quantifiable source of error in both years. Several other factors potentially contributing to the disparity between the two independent estimates of chum salmon abundance in both years are discussed.